
Cloud โ๏ธ AI ๐ง DevOps - CS grad @UNTsocial ๐
Texas, USA
Joined April 2024
- Tweets94
- Following110
- Followers32
- Likes351
DK retweeted
Whoa thatโs a full rainbow, all the way, double rainbow ohmygod, itโs a double rainbow all the way
DK retweeted
Full Course: MCP for Beginners
Build #AIAgents with #MCP (powerful framework for building secure, scalable, and real-world #AI apps). Explore core concepts, hands-on dev, advanced implementation, and how to build and deploy AI agents with MCP in #VSCode
youtu.be/VfZlglOWWZw?si=gw4-โฆ
DK retweeted
Learn how to supercharge #generativeAI workflows with NVIDIA DGX Cloud on #AWS & #AmazonBedrock Custom Model Import
๐ go.aws/44YoyQw
DK retweeted

Sundar Pichai@sundarpichai16 Jul 2025
United States
United StatesConnected via United States Android AppAccount-level information, not a live location or per-post device.
New advanced AI features coming to Search:
Gemini 2.5 Pro and Deep Search are rolling out in AI Mode to help with more complex questions and research - available now to Google AI Pro/Ultra subscribers. Plus, a new agentic capability directly in Search: AI-powered calling to local businesses, rolling out to all users in the US.
DK retweeted

AI at Meta@AIatMeta5 Apr 2025
United States
United StatesConnected via WebAccount-level information, not a live location or per-post device.
Today is the start of a new era of natively multimodal AI innovation.
Today, weโre introducing the first Llama 4 models: Llama 4 Scout and Llama 4 Maverick โ our most advanced models yet and the best in their class for multimodality.
Llama 4 Scout
โขย 17B-active-parameter model with 16 experts.
โขย Industry-leading context window of 10M tokens.
โขย Outperforms Gemma 3, Gemini 2.0 Flash-Lite and Mistral 3.1 across a broad range of widely accepted benchmarks.
Llama 4 Maverick
โขย 17B-active-parameter model with 128 experts.
โขย Best-in-class image grounding with the ability to align user prompts with relevant visual concepts and anchor model responses to regions in the image.
โข Outperforms GPT-4o and Gemini 2.0 Flash across a broad range of widely accepted benchmarks.
โข Achieves comparable results to DeepSeek v3 on reasoning and coding โ at half the active parameters.
โขย Unparalleled performance-to-cost ratio with a chat version scoring ELO of 1417 on LMArena.
These models are our best yet thanks to distillation from Llama 4 Behemoth, our most powerful model yet. Llama 4 Behemoth is still in training and is currently seeing results that outperform GPT-4.5, Claude Sonnet 3.7, and Gemini 2.0 Pro on STEM-focused benchmarks. Weโre excited to share more details about it even while itโs still in flight.
Read more about the first Llama 4 models, including training and benchmarks โก๏ธ go.fb.me/gmjohs
Download Llama 4 โก๏ธ go.fb.me/bwwhe9
DK retweeted
Llama 4 Herd is here!
It brings a lot of goodies, like MoE architecture and native multimodality, enabling developers to build personalized multimodal experiences.
With Day 0 support in vLLM, you can deploy Llama 4 with @vllm_project now!
Let's dig into it.
(a thread)
DK retweeted
Thrilled to bring Metaโs Llama 4 Scout and Maverick to Foundry today, as we continue to make Azure the platform of choice for the worldโs most advanced AI models. azure.microsoft.com/en-us/blโฆ
DK retweeted
๐๐ฃ๐ ๐๐ฎ๐๐ฒ๐๐ฎ๐ ๐๐ ๐๐ผ๐ฎ๐ฑ ๐๐ฎ๐น๐ฎ๐ป๐ฐ๐ฒ๐ฟ ๐๐ ๐ฅ๐ฒ๐๐ฒ๐ฟ๐๐ฒ ๐ฃ๐ฟ๐ผ๐
๐
When designing a system, we often need clarification about the roles of a Load Balancer and an API Gateway. Most of our resources focus on their implementation rather than real-life use cases.
๐น ๐๐ฃ๐ ๐๐ฎ๐๐ฒ๐๐ฎ๐ sits between a client and a group of backend services. It performs the function of a reverse proxy by accepting all application programming interface (API) calls, aggregating the different services needed to fulfill them, and returning the right outcome. ๐จ๐๐ฒ๐ฟ ๐ฎ๐๐๐ต๐ฒ๐ป๐๐ถ๐ฐ๐ฎ๐๐ถ๐ผ๐ป, ๐ฟ๐ฎ๐๐ฒ ๐น๐ถ๐บ๐ถ๐๐, ๐ฎ๐ป๐ฑ ๐๐๐ฎ๐๐ถ๐๐๐ถ๐ฐ๐ are typical duties that API gateways take care of on behalf of a system of API services. Also, the API gateway can handle faults (circuit breaker) and log and monitor.
๐น ๐๐ผ๐ฎ๐ฑ ๐๐ฎ๐น๐ฎ๐ป๐ฐ๐ฒ๐ฟ is a service that distributes incoming traffic across many servers or resources. Usually, we have two or more web servers on the backend, and it ๐ฑ๐ถ๐๐๐ฟ๐ถ๐ฏ๐๐๐ฒ๐ ๐ป๐ฒ๐๐๐ผ๐ฟ๐ธ ๐๐ฟ๐ฎ๐ณ๐ณ๐ถ๐ฐ ๐ฏ๐ฒ๐๐๐ฒ๐ฒ๐ป ๐๐ต๐ฒ๐บ. Its primary purpose is to use resources optimally. A more equal task allocation and increased capacity can enhance the system's responsiveness and reliability. There are three high-level load balancers: hardware-based, cloud-based, and software-based.
๐น ๐ฅ๐ฒ๐๐ฒ๐ฟ๐๐ฒ ๐ฃ๐ฟ๐ผ๐
๐ server resides in front of backend servers and transfers client requests to these servers. Reverse proxies are typically used to increase security, speed, and dependability. A reverse proxy receives a request from a client, forwards it to another server, and then returns it to the client, giving the impression that the first proxy server handled the request. These proxies ensure that users don't access the origin server directly, giving the web server anonymity. They are usually used for Load balancing, where we need to deal with incoming traffic flow so that we may distribute that traffic between multiple backend servers or use them for caching.
So, the main thing that differs these two is that an ๐๐ฃ๐ ๐๐ฎ๐๐ฒ๐๐ฎ๐ ๐ถ๐ ๐ณ๐ผ๐ฐ๐๐๐ฒ๐ฑ ๐ผ๐ป ๐ฟ๐ผ๐๐๐ถ๐ป๐ด ๐ฟ๐ฒ๐พ๐๐ฒ๐๐๐ to the appropriate service and it handles requests for APIs, while a ๐๐ผ๐ฎ๐ฑ ๐ฏ๐ฎ๐น๐ฎ๐ป๐ฐ๐ฒ๐ฟ ๐ถ๐ ๐ณ๐ผ๐ฐ๐๐๐ฒ๐ฑ ๐ผ๐ป ๐ฑ๐ถ๐๐๐ฟ๐ถ๐ฏ๐๐๐ถ๐ป๐ด ๐ฟ๐ฒ๐พ๐๐ฒ๐๐๐ ๐ฒ๐๐ฒ๐ป๐น๐ between a group of servers and handles requests that are sent to a single IP address, which works at protocol or socket level (TCP, HTTP).
Some ๐ฒ๐
๐ฎ๐บ๐ฝ๐น๐ฒ๐ of API Gateways are Amazon API Gateway, Ocelot (.NET-based), Tyk, or Apache APISIX, while Load Balancers are Azure Load Balancer, HAProxy, or Seesaw.
An example of reverse proxy services is ๐๐ฝ๐ฎ๐ฐ๐ต๐ฒ ๐ฃ๐ฟ๐ผ๐
๐, ๐ก๐ด๐ถ๐ป๐
๐ผ๐ฟ ๐๐๐ฆ with additional modules (Url Rewrite).
#technology #softwareengineering #programming
DK retweeted
RIP Google and LinkedIn!
Deepsearch in Grok 3 is the future of job hunting.
Hereโs a step~by~step guide ๐
(Bookmarkย forย later)
Kubernetes Security (where to focus)๐ฅ๐ฅ
1. RBAC
2. Network Policies
3. Secrets Manager
4. PSD & PSA
5. Image Security & Runtime Protection
DK retweeted
Challenges of Using Different IaC Tools and Role of Terraform
๐Using different Infrastructure as Code (IaC) tools for different cloud providers presents several challenges.
โก๏ธHere's a breakdown of the issues:
๐Vendor Lock-In and Script Incompatibility:
1.When an organization uses tools specific to a cloud provider, such as AWS CloudFormation Templates (CFT) for AWS or Azure Resource Manager for Azure, the scripts created for one cloud platform are not compatible with others.
2.This means that if the organization decides to switch cloud providers, all of the existing automation scripts must be rewritten for the new platform. This can be very time-consuming and require a lot of effort.
๐Increased Complexity in Hybrid Cloud Environments:
Many companies operate in a hybrid cloud model, utilizing multiple cloud platforms such as AWS and Azure. In this model, if different tools are used for each cloud, a DevOps engineer has to learn multiple tools to manage the infrastructure across these different cloud providers.
๐Need to Learn Multiple Tools:
The variety of tools for different platforms means that DevOps engineers need to learn multiple tools, such as AWS CFT, Azure Resource Manager, and Heat templates for Openstack, depending on the platforms the organization uses. This increases the learning curve and the complexity of managing infrastructure.
๐Difficult Migrations:
The need to rewrite scripts when migrating from one cloud provider to another is a major problem. The scripts written for one cloud provider can not be used on another, requiring significant effort to rewrite them.
โก๏ธTerraform addresses these issues by acting as a common interface that translates instructions into API calls for each specific cloud provider.
โก๏ธThis tool can manage infrastructure across different cloud providers using a single language and set of scripts.
๐ Choosing the Right AWS Database: RDS vs. DynamoDB vs. ElastiCache! ๐
๐ก Struggling to decide which database is best for your workload?
โ
Need structured SQL data with transactions? โ RDS
โ
Looking for high-speed, serverless NoSQL? โ DynamoDB
โ
Want ultra-fast caching to boost performance? โ ElastiCache
โก Each has its strengths - pick the right one for your needs! Letโs break it down. โฌ๏ธ
DK retweeted
this is how i learn gpu programming and CUDA to make it easier for me:
- the most important part is to really understand how the GPU is structured and the memory hierarchy within it i.e. global memory, constant memory, caches, shared memory, registers, and so on.
- secondly, stop thinking sequentially. i mean think about how the execution of your operation will look like in parallel. think about the data flow and the operations involved.
- CUDA relies on grid of blocks, and blocks of threads. understand the distinction between when to map your data to the threads, and when to map the threads to your data.
- before jumping to writing any kernels, take a pen and a paper and just start visualizing the threads (and blocks) of the kernel and what you want them to do.
- for CUDA, make sure you understand pointers and how you can play around with the indices to fetch/write data.
- really pay attention while reading the book below, and also watch the corresponding video. read the code of others to try and understand what's happening there.
- practice, practice, practice.
Amazon provides three primary options for running containers: ECS (Elastic Container Service), EKS (Elastic Kubernetes Service), and Fargate.
Let's understand them one by one ๐๐


























